Summary of Work: In this project we investigate the mechanisms of DNA replication fidelity in E. coli by a combination of in vivo and in vitro approaches. In vivo, we investigate the specificity of mutation in the E. coli lacI gene in strains affected in various aspects of replication fidelity. For example, analysis of sequenced lacI mutations in wild-type, mismatch repair defective mutL strains, and proofreading defective mutDmutL strains, has allowed estimates to be made for the efficiencies and specificities of in vivo base selection, exonucleolytic proofreading and DNA mismatch repair. In vitro, we have developed novel in vitro fidelity assays, again using the lacI gene as a target, allowing measurement of the fidelity properties of purified DNA polymerase III in its various assemblies, ranging from the isolated alpha subunit to the complete holoenzyme (HE). Interestingly, the fidelity behavior of polymerase III in vitro is quite different from that in in vivo. Specifically, DNA polymerase III (a subunit or HE) in vitro produces an abnormally high level of (-1) frameshift mutations. These data point to the existence of a novel fidelity system in vivo that prevents (-1) frameshift mutations. This system is currently investigated by isolating E. coli mutants defective in this process. As part of this effort we have isolated novel mutants of the dnaX gene (encoding the t subunit of HE) that specifically enhance frameshifts and transversion mutations. We have also demonstrated a mutator effect for an existing dnaN mutant, encoding the pol III processivity factor (sliding clamp). Finally, we have developed a system to measure for the first time the differences between leading and lagging strand replication on the E. coli chromosome.